AMD Instinct MI250X meets the minimum VRAM requirement for Q4 inference of mistralai/Mistral-Large-Instruct-2411. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
AMD Instinct MI250X can run mistralai/Mistral-Large-Instruct-2411 with Q4 quantization. At approximately 101 tokens/second, you can expect Excellent speed - conversational response times under 1 second.
You have 68GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 60GB | 128GB | 100.72 tok/s | ✅ Fits comfortably |
| Q8 | 120GB | 128GB | 69.21 tok/s | ✅ Fits comfortably |
| FP16 | 240GB | 128GB | 33.88 tok/s | ❌ Not recommended |